As much of the research material Apple has been publishing in recent months indicates, the company is investing heavily in all kinds of artificial intelligence technologies. Apple will announce its artificial intelligence strategy in June at WWDC, as part of iOS 18 and its other new versions of the operating system.
In the latest Power On newsletter, Mark Gurman says the iPhone’s new AI features are expected to work entirely with a large offline, on-device language model developed by Apple. You can expect Apple to tout the privacy and speed benefits of this approach.
9to5Mac previously found code references in iOS 17.4 that referenced an on-device model called “Ajax.” Apple is also working on server-hosted versions of Ajax.
The downside to on-device LLMs is that they cannot be as powerful as models running on huge server farms, with tens of billions of parameters and data that are continually updated.
However, Apple engineers can likely take advantage of the full vertical integration of its platforms, with software tailored to Apple silicon chips inside its devices, to make the most of an on-device focus. On-device models typically respond much faster than request traffic through a cloud service and also have the advantage of being able to work offline in locations with limited or no connectivity.
While on-device LLMs may not have the same rich built-in database of knowledge as something like ChatGPT for answering questions about all sorts of random trivia, they can be tuned to be very capable at many tasks. You can imagine that an on-device LLM could generate sophisticated automatic responses to Messages or improve the interpretation of many common Siri requests, for example.
It also fits perfectly with Apple’s strict privacy compliance. There is nothing wrong with processing all emails and text messages downloaded through an on-device model, as the data remains local.
The models on the device can also perform generative AI tasks, such as creating documents or images, based on prompts, to get a decent result. Apple still has the flexibility to partner with a company like Google to turn to something like Gemini on the server for certain tasks as well.
We’ll know for sure what Apple plans to do when it officially announces its AI strategy at WWDC. The keynote begins on June 10 and the company will present all the new software features coming to iPhone, iPad, Mac, Apple Watch, Apple TV, Vision Pro, and more.
Keynote USA News
For Latest Apple News. Follow @Keynote USA News on Twitter Or Google News.
